There are a number of reasons that Novell® Cluster ServicesTM might fail to start. The most common reasons are described below.
NDS and Time Synchronization Problems
The following screen is an example of what you see on the server console when there are NDS® problems or time synchronization is not functioning properly.
Wait for a few minutes to see if the problem is resolved. If, after the wait, CLSTRLIB.NLM still does not load, check NDS health and time synchronization.
Follow the instructions in TID 10060600 (http://support.novell.com/cgi-bin/search/searchtid.cgi?/10060600.htm )to perform an NDS health check, and verify that time is synchronized.
To force time synchronization, complete the following steps:
1. Load MONITOR on the server console.
2. Go to Server Parameters>Time>TIMESYNC
Time Source and then enter the IP address of first server in your NDS
tree. Usually the first server is configured as a Single time server and it has the Master NDS replica.
3. Go to Server Parameters>Time>TIMESYNC
Reset> and set the value to ON.
4. Go to Server Parameters>Time>TIMESYNC
Restart Flag>, and then set the value to ON.
LAN Communication Problems
The following screen is an example of what you see on the server console when there are LAN communications problems.
You might see messages similar to those above when the entire cluster comes up. This is normal. The nodes are competing to see which will become the master node.
If you see a lot of these messages when the cluster has already formed and new nodes are joining the cluster, there could be a LAN communication problem. In this case check all LAN hardware and drivers. See LAN/SAN Problems for more information.
Failure to Find SBD Partition
The following screen is an example of what you see on the server console when a node which is trying to join the cluster cannot find the SBD Partiton.
If there are other nodes in the cluster already running, it means this particular node is disconnected from the SAN or the part of the SAN contains the SBD partition.
Check all LAN hardware and drivers. See LAN/SAN Problems for more information.
Missing NDS Properties
The following screen is an example of what you see on the server console when there are missing NDS properties.
One of the cluster nodes has lost its association with the cluster container.
Go to each Server object in ConsoleOne®, click the Other properties tab and verify that the NCS:NetWare Cluster property is associated with the cluster container. You might have to click the Add button to see this property.
Too Many Characters in Load or Unload Script
The following screen is an example of what you see on the server console when a resource load or unload script contains more than 600 characters.
Reduce the number of characters in the resource load or unload script.
NDS Corruption or Too Many NICs or Bindings
A message similar to the following appears on your server console screen if NDS is corrupted or you have too many LAN cards or LAN card bindings:
CLSTRLIB - <FATAL> - <51>: ncslibLoadClusterConfig: ncslibResolveName failed ...
Repair NDS. See NDS and Time Synchronization Problems.
If you have too many LAN cards or LAN card bindings, remove multiple LAN card bindings and simplify your configuration. You can also apply the latest Cluster Services Support Pack, which resolves this issue.
This problem is resolved in NetWare Cluster Services version 1.01 SP2 and later.
A trademark symbol (®, TM, etc.) denotes a Novell trademark. An asterisk (*) denotes a third-party trademark. For information on trademarks, see Legal Notices.